Photo of Bexis

Today’s guest post is another tech-related discussion from Reed Smith‘s Jamie Lanphear. Given the increasing ubiquity of artificial intelligence (“AI”) in legal practice, the notion of AI prompts and output becoming yet another front in the never-ending ediscovery wars is concerning. Here are Jamie’s latest thoughts on the latest pertinent caselaw in this area. As always our guest posters deserve 100% of the credit (and any blame) for their efforts.

**********

The Blog has recently covered both the discoverability of AI prompts (here) and the Heppner decision out of the Southern District of New York (here). Since those posts, courts have continued to develop the law in this area — and some of the developments are encouraging. This post picks up where those earlier discussions left off, surveying the latest case law on privilege, work product, and waiver as they apply to AI prompts and outputs, and offering practical guidance for lawyers and in-house teams navigating an area where the rules are still being written.

Heppner: A Quick Recap and What It Leaves Open

As the Blog discussed here, in United States v. Heppner, 2026 WL 436479 (S.D.N.Y. Feb. 17, 2026), Judge Rakoff held that a criminal defendant’s communications with a generative AI platform were protected by neither attorney-client privilege nor the work product doctrine. The holding rested on several narrow facts: the defendant used a public AI tool on his own, without direction from counsel; the platform’s terms of service destroyed any expectation of confidentiality; and the materials were seized pursuant to a search warrant in a criminal case rather than sought through civil discovery. The court did not apply or analyze work product protection under Rule 26(b)(3) of the Federal Rules of Civil Procedure. The decision generated significant concern – but the limited application left a number of important questions unanswered.

Open Questions: Privilege in the Corporate AI Context

Heppner is what we have on privilege, but it does not tell us very much about how privilege will apply in the corporate context. The case involved a party acting on his own, not for the purpose of seeking legal advice from counsel. The in-house world presents meaningfully different scenarios.

Start with the threshold problem. Attorney-client privilege requires a communication between an attorney and a client, and when a corporate employee communicates with an AI chatbot, the communication is between the employee and a machine — not between the employee and counsel.

However, there are more-than plausible arguments that some of these interactions should remain protected. If an attorney directs an employee to use an in-house AI tool as part of the process of providing legal advice, the tool looks more like an agent than a third party — functionally similar to a paralegal or other intermediary acting at counsel’s direction. No court has endorsed that theory yet, but it is not an unreasonable application of existing doctrine to novel facts. Thus, in the first place, attorneys should think long and hard before directing clients (rather than their own offices) to do something like this.

The same logic applies in more routine scenarios. If an employee communicates with a colleague to gather information at counsel’s request for the purpose of providing legal advice to the company, those communications are privileged. What if the employee consults an AI tool instead of a colleague? The substance of the interaction is the same — the employee is gathering information counsel needs to provide legal advice. The only difference is the medium.

A related question arises with enterprise AI tools. If a company deploys an internal AI system and counsel interacts with that system to obtain information necessary to provide legal advice, there is a reasonable argument that the tool functions as the client’s agent. That scenario is not meaningfully different from an attorney communicating with a client’s accountant or other third-party agent to obtain information needed to render legal advice. Enterprise tools are deployed by the company, access-controlled, and integrated into the legal workflow. They are not consumer-grade chatbots with permissive terms of service. That factual distinction was critical to the outcome in Heppner.

None of these theories has been tested. But it is unlikely that courts will fail to find ways to adapt privilege doctrine to accommodate the realities of AI use. The law may lag science, but it does not ignore technology. The analogies are there, and courts have adapted privilege to new contexts before. The questions are not whether privilege will evolve, but how, and with what nuances.

Morgan: A Counter to Heppner

A recent decision out of the District of Colorado offers a notably different perspective on whether AI interactions are protected — at least under the work product doctrine. In Morgan v. V2X, Inc., No. 25-cv-01991-SKC-MDB, 2026 WL 864223 (D. Colo. Mar. 30, 2026), an employment discrimination case, a pro se plaintiff used AI in connection with litigation preparation. The defendant moved to compel disclosure of the AI tool the plaintiff was using and sought to amend the protective order to restrict AI use with confidential information. The discoverability of the plaintiff’s actual prompts and outputs was not directly at issue, but in working through the questions before it, the court engaged in an analysis that pushes back on several of the premises underlying Heppner—and the result is some useful support for the proposition that AI-assisted litigation materials are protected work product.

First, the court held that Rule 26(b)(3) broadly protects materials prepared in anticipation of litigation by a party, not merely by counsel. In civil cases, the rule’s plain language extends work product protection to parties and their representatives—a point that distinguished Morgan from Heppner, which was a criminal case governed by a different procedural rule.

Second, the court rejected the argument that using a third-party platform destroys confidentiality or waives protection. Drawing on Fourth Amendment case law, the court made a point that is both memorable and important: does anyone with a Gmail account forfeit all rights to confidentiality because Gmail has access to their emails? The court emphasized that in today’s world, nearly all electronic interactions pass through third-party systems, and courts are increasingly pushing back on the notion that using modern technology automatically destroys privilege.

Third, the court emphasized that work product (in contrast to attorney/client privilege) waiver requires disclosure to an adversary or under circumstances that substantially increase the likelihood that an adversary would obtain the information. Disclosure to an AI platform provider does not meet this standard—you are not disclosing information to your adversary, and there is no meaningful indication the information will end up in an adversary’s hands simply because it was entered into an AI system.

An important caveat: because the pro se litigant in Morgan was simultaneously the party and the advocate, Morgan does not resolve what happens when a non-attorney party — say, a corporate employee — uses AI independently, without attorney direction. In that scenario, the Heppner “gap” between party and counsel would exist even in the civil context. But the textual foundation of Rule 26(b)(3), as the Morgan court discussed it, supports an argument that such use would still be protected.

The bottom line is that while Morgan did not directly rule on the discoverability of prompts and outputs, it supports the position that AI-assisted litigation materials can fall within work product — and that the mere fact that information is processed or stored by an AI platform does not automatically result in waiver. This is a practical and sensible position that hopefully other courts will follow.

Tremblay: Prompts and Outputs as Opinion Work Product

As the Blog has noted (here), Tremblay v. OpenAI Inc., No. 23-cv-03223-AMO, 2024 WL 3748003 (N.D. Cal. Aug. 8, 2024), established that when counsel crafts AI prompts, both the prompts and the resulting outputs constitute opinion work product — the highest tier of protection — because the prompts reflect counsel’s mental impressions and opinions about how to interrogate the AI tool. Opinion work product is virtually undiscoverable absent a showing that counsel’s mental impressions are at issue and the need for the material is compelling.

This means that, at least until the case law further develops, having counsel create or direct the creation of AI prompts is not just a best practice. It has doctrinal consequences for the level of protection those materials receive. As a general principle, the closer an attorney is to the creation of the prompts, the stronger the argument that the materials are protected work product.

Waiver: Where Protection Breaks Down

Work product protection can be lost, and several recent decisions illustrate how that can happen in the context of AI use.

In Concord Music Group, Inc. v. Anthropic PBC, the court addressed waiver across multiple discovery orders over the course of 2025. See, e.g., 2025 WL 1482734 (N.D. Cal. May 23, 2025); 2025 WL 3677935 (N.D. Cal. Dec. 18, 2025). Consistent with Tremblay, the court held that the prompts and related settings used in the publishers’ investigations were opinion work product. The publishers voluntarily produced the prompts and outputs they relied on in their complaint and filings — roughly 5,000 prompt-output pairs. The defendant then sought broadly all of the remaining prompts and outputs, including those that did not support the publishers’ claims. The court denied that request as overbroad, holding that waiver must be closely tailored to the needs of the opposing party and limited to what is necessary to rectify any unfair advantage gained.

Later in the litigation, the publishers conducted an investigation into the effectiveness of the defendant’s AI guardrails — essentially testing whether the guardrails could be circumvented using various prompts. When they indicated they intended to put a witness on the stand to testify about that investigation — including the prompts the witness used and what he learned using them — the court found at-issue waiver under the sword-and-shield doctrine. In deciding where the draw the line, the court held that protection had been waived for the prompts that had been provided to the witness to conduct the investigation—including those provided to him by counsel—and the associated outputs, but those that remained with the attorney (and therefore were not placed at issue) remained protected. So here we see waiver of opinion work product, but a narrow tailoring of that waiver by the court.

A similar result played out in T.B. v. Big Brothers Big Sisters of N.Y.C., No. 452864/2021, 2025 WL 2443502 (N.Y. Sup. Ct. Aug. 21, 2025), where plaintiff’s counsel disclosed a portion of a ChatGPT transcript during a deposition. The court held that this placed the summary at issue and waived work product protection, and ordered production of the full AI exchange.

While these waiver principles are not new (or unique to AI), what AI potentially changes is the frequency with which the issue might arise. Tools that make it easy to generate analytical work product—evaluating complaint data, synthesizing the medical literature, identifying patterns in adverse event reports—also make it easy to rely on that work product in ways that place it at issue. The discipline and forethought are the same that have always been required. AI may just create more opportunities to get it wrong.

Where This Is Headed

As Judge Rakoff observed, AI “presents a new frontier in the ongoing dialogue between technology and the law.” That much is obvious. What is increasingly clear from the case law, however, is that courts are finding ways to apply existing doctrines to new technology. Work product protection extends to AI interactions. Attorney-crafted prompts are opinion work product and therefore receive the highest level of protection. And the mere use of a third-party AI platform does not automatically waive that protection.

The attorney/client privilege question is less settled, but the analogies are there. And there is no reason to think that courts will not adapt privilege doctrine to accommodate AI—just as they have adapted it to accommodate other evolutions in the provision of legal services. The waiver landscape requires attention, but the principles exist and the framework for applying them is well established. What remains is for courts to work through the application in this particular context.

Photo of Michelle Yeary

This post is from the non-Reed Smith side of the blog.

If you’re going to forum shop, you still need a legitimate forum. That was the problem for plaintiffs in Rodriguez v. Thoratec Corp., 2026 U.S. Dist. LEXIS 88964 (N.D. Cal. Apr. 20, 2026). Texas and Arkansas residents brought suit in California over an implanted medical device—even though the device was implanted in Texas, all treatment occurred in Texas, and the decedent died in Texas. The only apparent hook? One defendant’s alleged citizenship in California. That wasn’t enough. The court granted defendants’ motion to transfer to the Northern District of Texas.

The case involves a left ventricular assist device (LVAD) implanted in Texas. Following the recipient’s death, his family sued, alleging device-related complications. But there was no dispute about where the relevant events occurred: implantation, treatment, and death all took place in Texas.

Before turning to convenience—the question of whether the case should be transferred–the court first confirmed that the case could have been brought in Texas. It could. The manufacturer was subject to specific personal jurisdiction there because it sold the device in Texas, and plaintiffs’ claims “arise out of and relate to” those contacts. Any potential jurisdictional gaps as to other defendants disappeared when they consented to personal jurisdiction in the Northern District of Texas “for the limited purposes of the case.” Id. at *6-10. Diversity jurisdiction supplied subject matter jurisdiction. Id. at *10-12. And since it was undisputed that a “substantial part of the events … giving rise to the claim occurred” in Texas, venue is proper there.  Id. at *13.

That cleared the way for the familiar eight-factor transfer analysis under 28 U.S.C. § 1404(a). And nearly every factor pointed in the same direction–east.

Plaintiffs’ choice of forum? While ordinarily a heavy weight on the scale, here it was entitled to little deference because plaintiffs were not California residents. Id. at *13-14. Moreover, California had no meaningful connection to the operative facts. None. So, plaintiffs’ choice did not weigh against transfer. 

Convenience of the parties? Also Texas. Plaintiffs lived in Texas and its neighbor Arkansas. Defendants undercut plaintiffs’ only counter—concerns about securing company witnesses live at trial—by agreeing to produce appropriate witnesses in Texas.  Id. at *14-15. Another factor for transfer.

Convenience of the witnesses? This one is often “the most important factor” considered by courts, nd the primary concern is third-party witnesses. Here, that was decisive.  The key third-party witnesses—the implanting surgeon and treating physicians—were in Texas. To the exten it mattered, defendants’ concession addressed their own employees. Id. at *15-16. Transfer again.

Ease of access to evidence? Plaintiffs tried the old “documents are in California” argument. The court wasn’t buying it. In the age of electronic storage, the physical location of the PMA file means little. Id. at *16-17. Neutral.

Familiarity with the law? Texas law likely governs a Texas-based injury, and a Texas federal court is better positioned to apply it. Id. at *17. Transfer.

Local interest? Strong in Texas, where the device was used and the alleged injury occurred. California’s interest in regulating manufacturers within their borders exists but is “decidedly weaker” when the injury happens elsewhere. Id. at *18. Transfer.

Court congestion? This is just about reading the federal court statistics. The Northern District of Texas was less congested. Id. at *18-19. Transfer.

At the end of the day, this was not a close call. When the plaintiffs, the medical care, the alleged injury, and the key witnesses are all in Texas, filing in California looks less like a reasoned choice and more like a detour. Section 1404(a) exists to correct exactly that. Courts do not need to indulge choices with no meaningful connection to the dispute.

Photo of Bexis

Our PMA Preemption Scorecard now collects 729 opinions, the vast majority of which find preemption (which is why it’s a “scorecard”).  Only three of them are from the Land of Enchantment, New Mexico – with two of them being in the same case.  The third is Yusnukis v. Nevro Corp., 2026 WL 509227 (D.N.M. Feb. 24, 2026).  The complaint alleged manufacturing and warning (not design) defects, breach of express and implied warranties, a negligence claim resembling negligence per se, fraud and consumer fraud violations.  Par for the course in 2026, we’d say.

But the product was a PMA-approved spinal cord stimulator, so the entire kit and kaboodle was dismissed, largely on preemption grounds.  Also par for the course, given the many preemptive FDA requirements to which PMA devices are subject plus the FDCA’s prohibition against private enforcement/causes of action. “Where a plaintiff misses this narrow gap, the claim is preempted and subject to dismissal.” Id. at *3 (citation omitted).

The preemption discussion starts, as many do, with judicial notice of both the device’s PMA and subsequent supplemental PMAs.  Id. at *4.  Plaintiffs often try to ward off preemption, at least on the pleadings, by not alleging anything about the device’s FDA pedigree.  Thankfully, Yusnukis, like most courts, didn’t let them.  The decision also noticed the fact, but not the truth, of the device’s warnings.  Id.

As is apparent from our PMA Preemption Scorecard, the only typical product liability claim the ever gets much purchase in PMA preemption cases are allegations of manufacturing defect.  Not in Yusnukis.  Plaintiff claimed that the device’s “leads deviated from FDA-approved specifications by exhibiting impedance values greater than 18 Ohms.”  Id. (footnote omitted).  Seemingly out of nowhere pops something called a “connect module,” which plaintiff claimed the defendant “failed to disclose.”  That claim was wrecked on the implied preemption side of the “narrow gap.”  “[C]laims premised on nondisclosure or misrepresentation to the FDA are ‘fraud-on-the-FDA’ claims, which the Supreme Court has held are impliedly preempted,” because “[e]nforcement of FDCA compliance rests solely with the federal government, not private litigants.”  Id. at *6 (citation to you-know-what omitted).

Then plaintiff argued that the manufacturing claim was a “parallel” claim.  It was not, because the 18 Ohms allegation did not correspond to anything in the PMA.

[E]ven accepting Plaintiff’s factual allegations as true, the [complaint] does not plausibly allege:  what FDA-approved specification required impedance at or below 18 Ohms, [or] how Plaintiff’s device departed from that specification beyond a conclusory reference to impedance readings.

Id.  The purported manufacturing defect thus nowhere alleged any “deviation” from any FDA requirement – let alone causation.  Id.  No deviation = no parallel claim = express preemption.

Next to go were the implied warranty claims.  “Courts consistently hold that implied warranty claims on dangerousness or ineffectiveness are preempted because they would require altering a PMA-approved design or labeling contrary to federal law.“ Id. at *7 (citations omitted).  Yusnukis was no different, as the merchantability claim asserted generally that the device was “unreasonably dangerous” and “not fit for its ordinary purpose.”  Id.  And once again, the complaint “do[es] not identify any FDA-approved specification requiring impedance below 18 Ohms.”  Buh-bye.  For some reason the plaintiff also alleged breach of the implied warranty of fitness for a particular purpose.  Another “par for the course” dismissal:  “The [complaint] does not allege that [defendant] knew of any particular purpose beyond the device’s ordinary indication.”  Id. at *8.  Nor did plaintiff plead reliance.  Id. at *9.  Go away, without even having to reach preemption.

Failure to warn also failed to work.  True, the warnings did not include the specifics of plaintiff’s alleged injuries, but that’s not a defense to express preemption (“different from or in addition to”).  Id. 

Plaintiff has not identified any FDA requirement mandating warnings about [plaintiff’s claimed injuries]. The FDA-approved labeling for [the] device contains warnings about [various other risks] . . . but it does not include the risks Plaintiff alleges. . . .  Plaintiff thus seeks to impose duties “different from, or in addition to” federal requirements, and the claim is expressly preempted under §360k(a).  Allowing such a claim to proceed would invite precisely the kind of state-law interference with FDA labeling determinations that Congress sought to bar, creating obligations that could undermine medical device regulation.

Id. at *10 (following Caplinger v. Medtronic, Inc., 784 F.3d 1335 (10th Cir. 2015) (which we discussed here) (other citations omitted).

Pursuing liability despite the device having the warnings that the FDA approved, plaintiff claimed that those warnings were nonetheless “false or misleading in any particular.”  Id. (quoting 21 U.S.C. §352).  No.  That section “govern[s] labeling in a general sense,” but plaintiff’s “state law claims swept far more broadly,” thus “the scope of the plaintiff’s state law claims exceeded the scope of the federal provisions she cited, they were not truly parallel.”  2026 WL 509227, at *10.  “[T]he state-law duty [plaintiff] identifies − to provide adequate warnings of foreseeable risks to a learned intermediary − sweeps more broadly than §352 and would require warnings about risks the FDA did not require.”  Id.  Express preemption precludes any “sweeping state duty to warn of foreseeable risks, without pointing to any specific FDA requirement requiring the warnings [plaintiff] seeks.”  Id.  Yusnukis thus provides a pithy dismissal of a bogus “misbranding“ warning argument that we see far too frequently.

Nor did “negligence” – actually negligence per se without using that phrase – fit through the narrow gap.  Plaintiff alleged a laundry list of compliance procedures that supposedly violated an equally long list of FDA good manufacturing practices.  Id. at *11.  The two lists, however, nowhere explained defect or causation.

Plaintiff fails to connect any specific regulatory violation to his injuries.  Plaintiff recites regulatory provisions and asserts proximate cause, but he offers no factual details showing how any specific violation resulted in . . . the basis of his injuries. Under [TwIqbal], such formulaic recitations are insufficient to state a plausible claim. . . .  [C]onclusory recitations of regulatory duties and proximate cause fail where the complaint does not allege facts connecting any violation to the plaintiff’s injuries.

Id. at 11 (citations omitted).  The claim thus failed on pleading grounds, and the court did not reach preemption.  Id.

The plaintiff in Yusnukis alleged five purportedly material misrepresentations as “fraud.”  Id. at *12.  Four of them failed because they were merely “forward-looking assurances rather than actionable misrepresentations of present fact.”  Id.  The fifth statement was closer to a present fact, but still inadequate, because plaintiff forgot the most basic element of fraud – falsity.  Id. at *13.  Further, none of these allegations pleaded either intent or causation with the necessary particularity.  Id. at *12-13.

Express warranty was also dismissed.  Boilerplate “safety and effectiveness” would-be warranties failed practically every element and were dismissed.  Id. at *14.  Purported “future performance” warranties were also boilerplate:

Plaintiff alleges only that the device was “not reasonably safe,” caused “numerous serious side effects,” resulted in “severe and permanent injury,” and “did not deliver the claimed results.”  These are generalized assertions of product failure, not factual allegations that the device failed to conform to any particular affirmation, promise, or description that formed the basis of the bargain.

Id.

Last and least were plaintiff’s supposed consumer protection claim.  Those allegations were “little more than a barebones recitation of the elements of a [New Mexico consumer protection] claim.”  Id. at *15.  Even worse, by the end of the complaint, plaintiff had given up even pretending to plead facts and instead “relie[d] on incorporation by reference, leaving the Court to piece together the claim.”  Id.  That failed even Rule 8’s ordinary pleading standards.  Id.  “[R]eliance on incorporation by reference and recitation of statutory elements does not cure the absence of factual detail linking a particular representation to the sale of goods.”  Id.

Seven claims up; seven claims down.  Yusnukis is a comprehensive rejection of most of the usual medical device product liability allegations, on both preemption and pleading grounds.  Any defendant litigating a PMA medical device case in New Mexico should be aware of it.

Photo of Bexis

Early Bird pricing may be history, but the good news is the 2026 Drug and Medical Device Seminar is shaping up to be anything but ordinary. If you like smart people, sharp ideas, and a little Boston energy, you’re in the right place.

What’s New and Big This Year

From May 20–22 at The Westin Copley Place in Boston’s bustling Back Bay, you’ll dig into the topics everyone in the field is talking about:

  • The newest twists on Federal Rule of Evidence 702 and expert challenges
  • Global mass tort coordination and cross‑border strategy
  • Hot‑button litigation trends like PFAS, ethylene oxide, and the push toward “clean pharma”
  • Direct insights from medical device GCs, in‑house leaders, and regulatory voices who don’t sugarcoat anything

Plus, this year’s program adds fresh sessions on AI in litigation, jury persuasion, MDL strategy, and science‑driven defense themes. DRI’s programming is all designed to give you takeaways you can use the moment you’re back at your desk.

Networking That Actually Matters

This seminar draws the people who move the needle in drug and device defense. You’ll meet trial lawyers, in‑house counsel, scientists, and industry leaders tackling the same challenges you are. And with attendees from companies like AbbVie, Becton Dickinson, GSK, Medtronic, Sanofi, and Stryker, the conversations you have here can genuinely shift your practice.

Boston. May. Your Next Big Professional Boost.

Prices are already up now that Early Bird has closed, but you can still save up to $500. If you want in, lock it in.

Register here: https://bit.ly/4ssfwFL

Photo of Eric Hudson

This post is not from the Reed Smith or Dechert sides of the blog.

We previously posted about the Delaware Supreme Court’s ruling that proper application of Delaware Rule 702 required the exclusion of plaintiffs’ general causation experts in the Zantac litigation. Since then, we’ve been waiting for the other shoe to drop.  Last week, it dropped with a resounding boom. Today’s decision applied the Delaware Supreme Court’s ruling and dismissed over 80,000 cases. In re Zantac (Ranitidine) Litig., No. N22C-090101, 2026 WL 1009008 (Del. Super. Apr. 14, 2026).

The decision focuses on the question of which plaintiffs are bound by the exclusion of the plaintiffs’ general causation experts. The court denied plaintiffs leave to supplement their expert reports on December 1, 2025, and the defense argued that all plaintiffs with cases filed before December 1, 2025, are bound by the ruling. Recognizing that the litigation was about to be eviscerated, plaintiffs made a number of desperate arguments.

Continue Reading No Mulligans in Delaware – Exclusion of Plaintiffs’ General Causation Experts Results in Dismissal of 80,000 Zantac Cases
Photo of Lisa Baird

Can we all agree that as between vaccines and [checks notes] “snort[ing] cocaine off of toilet seats”, it is the vaccines that should make us “not scared of a germ”?

Ok, maybe the [checks notes] Secretary of Health and Human Services wouldn’t agree, but c’mon—the evidence that vaccines save lives is overwhelming. 

As Perry v. Marteney, __ F.4th __, 2026 U.S. App. LEXIS 10101, 2026 WL 946152 (4th Cir. Apr. 8, 2026) recounts:

  • Smallpox went from killing 10% of London’s population in the 17th century and 500 million people globally between 1880 and 1980 to completely eradicated by 1980 after a global vaccination campaign.
  • Deaths caused by [other diseases] have also been significantly reduced since vaccines were first developed: diphtheria (100%); polio (100%); measles (100%); mumps (100%); rubella (100%); whooping cough (99.3%); tetanus (99.2%); chickenpox (81.9%); and hepatitis B (80.2%).

In fact, one of the amazing things about vaccines is that not only do they provide disease protection for the person vaccinated, they also can provide protection for the people around them, including the medically fragile whose conditions won’t allow them to get vaccines themselves. 

Or at least they can provide “herd immunity” if a large enough percentage of the herd is vaccinated.  Vaccine mandates work well to accomplish that goal, but they get a lot of push-back, including by those with religious objections.

Perry addressed the collision between compulsory vaccination laws and asserted religious rights.  In a divided opinion, the Fourth Circuit reversed a preliminary injunction that had allowed an unvaccinated child to remain enrolled in an online West Virginia public school based on her parents’ religious objections to vaccination.  The majority opinion reaffirms the continuing force of Jacobson v. Massachusetts, 197 U.S. 11 (1905), and falls in line with other opinions we have written about as of late. 

Like Jacobson, the Perry majority acknowledges the breadth of state police power when public health is at issue, and the limited role the First Amendment’s Free Exercise Clause plays where a neutral, generally applicable vaccination mandate is involved.  The dissent, by contrast, reads recent Supreme Court decisions as significantly expanding strict scrutiny protection for parental religious objections in the public-school setting and would have upheld the injunction.

The West Virginia law in question requires children attending “public, private or parochial” schools to be immunized against a list of serious infectious diseases, including chickenpox, hepatitis B, measles, meningitis, mumps, diphtheria, polio, rubella, tetanus, and whooping cough.  W. Va. Code § 16‑3‑4(c).  

All 50 states have some form of compulsory vaccination requirement, but West Virginia is one of the few that provides no religious exemption.  West Virginia’s law only allows for medical exemptions, when “immunization is contraindicated or there exists a specific precaution to a particular vaccine,” as certified by a licensed physician.  W. Va. Code § 16‑3‑4(h)(1).  Even then, West Virginia doesn’t just take the doctor’s note at face value—a state Immunization Officer decides whether there is “sufficient medical evidence” to justify the medical exemption, and that decision is then subject to appeal to the State Health Officer, and judicial review under the state Administrative Procedures Act.

At the same time, West Virginia requires all children to attend school, save for those taught through approved home instruction, “learning pods” or “microschools”.  These children are not subject to the vaccination requirement because they do not attend “schools” within the meaning of the vaccination mandate.

Perry involved a child who was dis-enrolled from a state-run public virtual academy (where the online content delivered by state-certified teachers) because she was not fully vaccinated as the statute required.  After her parents failed to obtain a religious exemption from the vaccine requirement, they sued and won a preliminary injunction against enforcement of the mandate.

The Fourth Circuit then reversed.  It began by situating West Virginia’s law within a long line of cases recognizing broad state authority to enact health measures, including vaccination, starting with Jacobson which had upheld a smallpox vaccine mandate while warning against trying to uphold public health if “each [man] is a law unto himself.”

The long-standing legal test is that a neutral and generally applicable law that incidentally burdens religious practice is subject only to rational-basis review, whereas governmental burdens on religious practice that are not neutral and generally applicable are subject to strict scrutiny.  Notably, Employment Division, Department of Human Resources of Oregon v. Smith, 494 U.S. 872 (1990), used compulsory vaccination as a paradigmatic example of a neutral and generally applicable, legally-required “civic obligation” that could not be made contingent on constitutionally required religious exemption without effectively turning every believer into “a law unto himself.”

Neither side really argued that West Virginia’s vaccination statute was not neutral, as the law neither targets nor disfavors religion on its face or in its operation. 

But as to general applicability, the question was whether the allowance for medical exemptions and the statutory classifications (that it applied to school students but not adult school workers; and it applied to virtual school students but not homeschoolers) made the vaccine mandate not generally applicable. 

The court found it notable that West Virginia officials have no authority to grant ad hoc exceptions for religious or other reasons.  The only exemptions allowed are medical, and those require “sufficient medical evidence” that the immunization is contraindicated or subject to a specific precaution for a particular child.  Moreover, that limited statutory exemption was constrained by appellate review options.  The majority thus fell in line with every other circuit (the First, Second, Third, and Ninth Circuits) to have addressed vaccine mandate medical exemptions, agreeing that they do not undermine the general applicability of a vaccine law

As to the distinctions drawn between populations subject to the mandate, the Fourth Circuit concluded they also did not deprive the mandate of its general applicability.  Differential treatment between secular categories (like between virtual school students and homeschoolers) does not raise the suspicion of religious animus that distinctions between religious and secular conduct would.

With the finding that the law was neutral and generally applicable, the rational basis standard was readily met:  The state’s interest in preventing the spread and severity of serious infectious diseases among schoolchildren was legitimate if not compelling, and compulsory vaccination is crafted to address that concern.  Moreover, the criticisms hurled at the law did not make it irrational.  For example, while both virtual students and homeschoolers may pose less risk of spreading disease, virtual students can and do participate in in-person school activities and thus had a different risk profile than those homeschooled. 

In fact, the Perry majority concluded that the state’s interest was so strong, and its policy so on point, that the law would pass strict scrutiny if that were the test to be applied, as the dissent argued.   

The dissent not only wanted strict scrutiny, it argued that the case should be evaluated not through the Jacobson framework, but rather under the framework from Wisconsin v. Yoder, 406 U.S. 205 (1972) and Mahmoud v. Taylor, 606 U.S. 522 (2025).  Under Mahmoud in particular, a law that “substantially interferes with the religious development” of a child and the “critical right of parents to guide the religious development of their children” is subject to strict scrutiny “regardless of whether the law is neutral or generally applicable.”

But unlike Yoder and Mahmoud which would have required children to sit through lessons (formal or otherwise) that were in conflict with their parents’ religious beliefs, West Virginia’s vaccination law is a public health measure, not an form of ideological or value-based instruction or messaging.  The majority also found it notable that nothing in Yoder or Mahmoud purported to overrule, or even undermine, Jacobson, and the lower courts are obligated to follow those precedents unless and until the Supreme Court says otherwise.  Here’s to hoping that day never comes. 

But at least for now, in the Fourth Circuit, state vaccination mandates that (1) are facially neutral, (2) provide only medically grounded exemptions with constrained professional judgment and a review procedure, and (3) apply across the public school system will be evaluated under rational-basis review and are likely to be upheld, notwithstanding parental religious objections.

Photo of Stephen McConnell

That title might be overblown, because we are discussing only two cases.  But one of them is the Roundup case, and we could not resist the cheesy wordplay. 

Roundup is neither a drug nor device. It is regulated by the Environmental Protection Agency (EPA), not the Food and Drug Administration (FDA). Nevertheless, the Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA) will feel familiar to our readers, because it preempts any state law “requirements for labeling or packaging in addition to or different from those required under this subchapter.”  That express preemption did not stop a Missouri jury from concluding that Roundup labeling was inadequate because it did not warn of the risk of cancer.  But such a label would have contravened the EPA’s repeated conclusions that Roundup’s key ingredient, glyphosate, did not cause cancer. The jury verdict was a miscarriage of justice because the warning label demanded by the plaintiff would have made the product misbranded.  The product could not comply with both the federal and state law requirements. That impossibility means that the plaintiff’s claim should have been impliedly preempted as well as expressly preempted.  

The appellant brief is splendidly written. That is hardly a surprise, especially when we see the names of Paul Clement and Mike Imbroscio on the cover.  You should go on to the SCOTUS blog and read the brief.  See here. Among the many things that it does remarkably well, it takes a tour of the regulatory record and makes a compelling case that the EPA arrived at a sound scientific conclusion that glyphosate does not cause cancer, and that a contrary announcement by a working group of the International Agency for Research on Cancer (IARC) — the main basis for the plaintiff’s claim — was bunk. 

As you would expect, many amicus briefs have been filed in this matter.  Most, sadly, side with the plaintiff.  But the one that counts the most was filed by the U.S. Solicitor General (sometimes referred to as the Tenth Justice) and that brief comes down forcefully in favor of express and implied preemption.  The SG also pounded home the concern that a cacophony of different state jury decisions will cause chaos: “Lost in that noise: EPA’s considered judgments about what warnings are actually necessary to protect public health, and any hope of uniformity.” 

We do not mean to suggest that the respondent brief was written by slouches. (We clerked for Ninth Circuit Judge William Norris, and he told us that when certiorari was granted for any of his decisions, he always was astounded by how wonderful the briefing was at the SCOTUS level.) The Keller Postman firm, along with others, is onboard, and they have ample intellectual wattage. Their brief is clever, thorough, and wrong.  They emphasize that FIFRA provides that the product’s registration is not a defense.  They argue that Roundup was misbranded because it was unsafe. Thus, according to their account, the plaintiff was seeking to vindicate FIFRA, not contradict it.  The respondent asserts that the defendant faced no impossibility.  If it could not change the label to match the IARC view, it could have simply stopped selling the product. 

Remarkably, the respondent also tried to exploit the SCOTUS decision in Loper Bright, suggesting it means that courts should not defer to the EPA’s determination that glyphosate does not cause cancer. (Long ago, we predicted this potential distortion of Loper by the plaintiffs’ bar.) Finally, the respondent argued that the defendant waived its arguments by failing to seek a Bates jury instruction at trial on the relevant FIFRA misbranding instructions.  Along the way, the respondent brief cites Marbury and Carolene Products in ways that seem, frankly, bizarre. 

To our admittedly biased eyes, the appellant briefing wins the day on all these arguments.  Preemptive force must be given to rules “under” FIFRA, which would cover the EPA’s determination that glyphosate does not cause cancer. The respondent wants to turn that “under” into a “by” and suggests there is no difference.  But Clement et al. point out that “There is no preposition exception to textualism.” The appellant also makes clear that the respondent’s stop-selling solution runs afoul of Bartlett. And one has to marvel at the temerity of the respondent to invoke Loper, given that Clement is the lawyer who won that case. The first rejoinder on the Loper issue is that it was not raised below, nor was it raised by any of the many amicus briefs. The respondent says that “is not the result of a collective failure to issue spot.” Second, Loper held that courts should not defer to agency statutory interpretation, but such statutory interpretation was not at issue here.  The FIFRA preemption provision is as clear as can be, and no one suggests the EPA acted outside its statutory authority. “Loper may change a lot, but it has zero relevance for that straightforward construction of FIFRA’s  unambiguous statutory text.”  That, friends, is a smack down.  

As for the procedural waiver argument, that was never raised by the plaintiff in the Missouri appellate courts, thereby “forfeiting any forfeiture.”  The last word – the ending of the appellant’s reply – is elegant: “FIFRA seeks balance and uniformity. The plaintiff’s bar seeks something altogether different, without regard to whether it drives products that farmers need from the market. The Court should vindicate Congress’ judgment.” 

The oral argument is on April 27. Bring your popcorn. 

We also want to alert our readers to certiorari briefing in BNSF Railway Co. v. Tanner Lynn.  The petition was filed by the great Lisa Blatt and her team at Williams & Connolly.  There is also an excellent amicus brief authored by the Washington Legal Foundation. The question presented is: “Whether the Commerce Clause permits a State to extract consent to jurisdiction from nonresident corporations as a condition of their right to do business in the State, even for cases involving nonresident plaintiffs and out-of-state conduct.” The BNSF cert petition invites the High Court to follow up on Justice Alito’s suggestion in the execrable Mallory decision that even if jurisdictional consent via corporate registration does not violate due process, it violates the dormant Commerce Clause.  (This blog has addressed the dormant Commerce Clause issue.)


Few cases caused such immediate mischief as Mallory did.  Litigation tourism, which had been reeled in by the Bauman and BMS cases, received a new lease on life from Mallory.  The cert petition pieces together the various decisions in Mallory to show that five justices seemed receptive to the dormant Commerce Clause argument. As has been said more than once, the key to SCOTUS litigation is being able to count to five. States like Pennsylvania in Mallory, or Minnesota in BNSF, have “no legitimate local interest” in “adjudicating a foreign-cubed lawsuit.”

We very much hope that SCOTUS grants cert for the BNSF case. It is the right case, the right time, and the right advocate. 

Photo of Michelle Yeary

Benzene-based consumer class actions are becoming almost routine. So are their dismissals on preemption grounds. It takes something a little different to make one stand out. The plaintiff in Leonard v. CVS Pharmacy, 2026 WL 986311 (N.D. Cal. Apr. 13, 2026), tried just that, invoking a res ipsa loquitur–type theory to plead “parallel” CGMP violations. That move may sound familiar to those who follow medical device litigation, where it has been tried—and rejected—before. But as far as we know, this is the first time plaintiffs have attempted to import that playbook into the growing wave of benzene cases.

It didn’t work here either.

We probably shouldn’t be surprised. This was already plaintiff’s second attempt to bring her consumer protection claims against the manufacturers of generic Mucinex. The court previously dismissed her claims as expressly preempted after concluding that the FDA’s approval of the drug’s ANDA, which disclosed that its manufacturing process involved a carbomer that could introduce benzene, necessarily determined the product complied with the FDCA. Claims that the product was “adulterated” or “misbranded” because it contained benzene imposed requirements different from federal law and were therefore preempted. Id. at *1. Plaintiff was given leave to amend. She did—but not successfully.

Round two swaps out express preemption arguments for a “parallel violation” theory grounded in CGMPs. Plaintiff alleged that defendants violated those regulations and that the violations resulted in benzene contamination. But as the court quickly recognized, the theory had a glaring problem: it was entirely unsupported by facts. Indeed, “the only fact Plaintiff alleges to support her allegations that Defendants violated the cGMPs, is the presence of benzene itself.” Id. at *2. But that’s just circular reasoning.

Start with the basics. CGMPs do not regulate the presence of benzene. In fact, the FDA approved the product with the condition that it could contain benzene below a certain threshold (2 ppm). Id. at *3. So, the mere presence of benzene does not establish a CGMP violation. And if plaintiff’s position is that any benzene violates CGMPs, that would put her theory squarely at odds with the FDA’s approval—creating the very conflict that preemption forbids.

So, plaintiff tried a different tack suggesting that maybe the levels were too high. But here again, the complaint comes up empty. There are no factual allegations about the benzene levels in the product plaintiff actually purchased—only the assertion that benzene levels can be “astronomically high.” Id. But saying something can happen is not the same as alleging that it did happen. Saying a product can contain high levels of benzene is like saying a lottery ticket can be a winner. True—but most aren’t. Plaintiff needs “more than a sheer possibility” to satisfy TwIqbal. She didn’t get there.

Plaintiff also attempted to anchor her claims by listing various CGMPs—written procedures, laboratory controls, recordkeeping obligations. But again, the complaint offered no facts explaining how any of those requirements were actually violated. A laundry list of regulatory provisions, untethered to factual allegations, likewise does not satisfy federal pleadings standards. Again, plaintiff relied exclusively on the presence of benzene as purported proof that defendants violated the CGMPs. But the court was unwilling to take the leap plaintiff invited–to infer that the failure to keep written procedures, laboratory controls, or laboratory records caused the presence of benzene. Id.

That’s where the res ipsa theory ultimately collapses. The premise is that benzene should not be present unless something went wrong, so the court should infer a CGMP violation. But as courts have already recognized in the device context, complex medical products and complex regulatory schemes do not lend themselves to that kind of shortcut. Without factual allegations connecting the dots, there is no “parallel” claim—just speculation dressed up as doctrine.

This time, because plaintiff advanced a new theory, she gets one more chance to try to plead it. But stepping back, the broader takeaway is becoming hard to ignore. Plaintiffs keep looking for a way into these benzene class actions—and keep running headfirst into preemption. Express preemption has blocked one path. Implied preemption has closed off another. And now, even repackaging the claim as a res ipsa-style “parallel violation” theory meets the same fate.

At some point, the pattern matters. Plaintiff may get another opportunity to re-plead in this case, but it is difficult to see what avenue remains. Without facts showing an actual, non-preempted violation tied to her product, simply relabeling the theory is not going to get her over the hump. So far, every path leads back to the same place: dismissal.

Photo of Bexis

The learned intermediary rule – that prescription medical product warnings are directed to (and written to be understood by) the professional medical personnel who prescribe them – is as close to unanimous as any tort doctrine that we know of.  We keep a running tab on the 50 states, and the rule has support in every state, including 37 highest courts (or statutory adoption).  The learned intermediary rule is so widespread because the policies supporting it are likewise universally recognized.  There are several such supportive policies, but one of the most important is to preserve the institution of the physician-patient relationship.

Direct-to-patient warnings from manufacturers and others in the chain of distribution of prescription medical products would inevitably complicate and disrupt the physician-patient relationship.  Unlike physicians, who tailor their informed consent process to the medical needs of particular patients, remote manufacturers could only provide generalized information – and none at all about off-label uses.  If the law were to require persons other than prescribing physicians to send generalized information directly to patients (assuming that it is even practical to do so, another reason for the rule), that information would inevitably conflict with the patient-specific medical advice that the prescribers themselves are legally and ethically required to provide their patients.

Again, that conflict would inevitably occur with off-label uses, given FDA restrictions on what regulated entities are required to say.  Off-label use remains a significant part of medical practice.  In his 2021 off-label use law review article Bexis reported:

More recent estimates indicate the prevalence of off-label use remains notoriously high among some patient populations.  An estimated twenty-one to fifty percent of all prescriptions are for off-label indications and in some patient groups, this number may exceed eighty percent.  Off-label use is particularly prevalent in oncology and in pediatrics.

James Beck, “Off-Label Use in the Twenty-First Century: Most Myths & Misconceptions Mitigated,” 54 UIC J. Marshall L. Rev. 1, 25-26 (2021) (footnotes and quotation marks omitted).  Two years later a Congressional Research Service report reached essentially the same conclusion. 

In a 2006 study of drug prescribing by office-based physicians, 21% of prescriptions were written for off-label uses.  Of those off-label prescriptions, the study’s authors found that 27% were backed by strong scientific support. . . .  An econometric model from the National Ambulatory Medical Care Survey estimated a 38% rate of off-label use.

Research has shown that more than half of oncology drug use is off-label.  A 2018 study examined 43 FDA-approved cancer drugs and compared their 99 labeled uses with the acceptable uses published by a national compendium Medicare relies on to make coverage decisions.  Of the 451 compendium-accepted uses, 56% were off-label. Of the off-label uses, the authors deemed 91% as “well-accepted off-label use.”

Hassan Sheikh, “Off-Label Use of Prescription Drugs,” at 2-3 (2023) (Canadian information and footnotes omitted).  That’s a lot of physician prescriptions that, but for the learned intermediary rule, would be disrupted by extraneous, but legally required information from FDA-regulated product manufacturers.  And the disruption to the physician-patient relationship that the learned intermediary rule prevents goes far beyond just off-label use situations.

Thus, it’s hardly a surprise that courts have repeatedly referenced damage to the physician-patient relationship as a basis for the learned intermediary rule.  Washington adopted the learned intermediary “primarily for public policy reasons focused on preserving the physician-patient relationship,” Rublee v. Carrier Corp., 428 P.3d 1207, 1217 (Wash. 2018), and that state’s highest court discussed the need for the rule to protect the physician-patient relationship at length in Dearinger v. Eli Lilly & Co., 510 P.3d 326 (Wash. 2022):

Physicians comprehend this complex information in a way the average lay person cannot.  Indeed, the FDA has recognized the information on the prescribers’ label is of “questionable” value when provided directly to patients and “relatively inaccessible to consumers. . . .”  [P]hysicians can give personally tailored warnings to patients in a way manufacturers cannot.  A physician can personalize warnings to a patient based on that patient’s medical history and needs.  “[I]t is only the physician who can relate the propensities of the drug to the physical idiosyncrasies of the patient.”  “A physician possesses the medical training to assess adverse health effects of a medical product and to tailor that assessment to a particular patient.”  Conversely, drug manufacturers cannot create individualized warnings because they do not know consumers’ medical information.  Thus, manufacturers issue broad, complex warnings that must be simplified by a learned intermediary − the physician − before being given to patients.

Dearinger v. Eli Lilly & Co., 510 P.3d 326 (Wash. 2022) (quoting McKee v. American Home Products, Corp., 782 P.2d 1045, 1051 (Wash. 1989), and Ruiz-Guzman v. Amvac Chemical Corp., 7 P.3d 795, 802 (Wash. 2000)) (regulatory citation omitted).  McKee also observed that mandating manufacturer warnings “may confuse and frighten the patient.”  782 P.2d at 1055.

Another state high court discussed the learned intermediary rule’s function to protect the physician-patient relationship extensively in Centocor, Inc. v. Hamilton, 372 S.W.3d 140 (Tex. 2012):

Despite the intricate web of modern healthcare providers and treatments, the bedrock of our healthcare system is the physician-patient relationship, and the ultimate decision for any treatment rests with the prescribing physician and the patient.  As a matter of both necessity and practicality, the duty to warn the patient of the potential risks and possible alternatives to any prescribed course of action rests with the prescribing physician. . . .  [Generalized information from a manufacturer] cannot explain the complex intricacies of the human body or supplant the detailed, interconnected nature of the practice of medicine, which necessitates that an informed intermediary help determine the best course of treatment for a patient’s particular symptoms.  To hold that each healthcare provider owes a separate and individual duty to warn each patient of all possible risks associated with a treatment prescribed by any doctor would . . . undermine the prescribing doctor’s physician-patient relationship. . . .  In most prescription drug contexts, the learned intermediary doctrine applies and the duty to warn the patient rests solely with the prescribing physician.

Id. at 166-67 (citations and footnote omitted).

“[I]mposition of a duty to warn the user directly would interfere with the relationship between the doctor and the patient.”  West v. Searle & Co., 806 S.W.2d 608, 613 (Ark. 1991) (citation omitted).  Plainly, “the highly personal doctor-patient relationship” is one of the “safeguards in place under the learned intermediary doctrine.”  Vitanza v. Upjohn Co., 778 A.2d 829, 845 (Conn. 2001).

When a patient consults with a physician seeking a prescription drug or restricted device, the patient also expects the physician to use his informed independent judgment to advise the patient and to prescribe the most appropriate use of the drug or device, based on his professional judgment.

Lacy v. G.D. Searle & Co., 567 A.2d 398, 400 (Del. 1989).  The “reasons” for the learned intermediary rule are “obvious” – patients “place primary reliance upon [physician] judgment,” and necessarily “[t]he physician decides what facts should be told to the patient.”  Wooderson v. Ortho Pharmaceutical Corp., 681 P.2d 1038, 1052 (Kan. 1984).  Only doctors know exactly how a drug is being prescribed, and can tailor the informed consent discussion to the relevant risks:

[I]mposing a duty to warn upon the manufacturer would unduly interfere with the physician-patient relationship. . . .  []Since the typical manufacturer’s warning provides a list with scores of potential side effects, no matter how minute the possibility of occurrence, the lay consumer might overreact to such warnings and forego beneficial, or even vital, medical treatment.

Larkin v. Pfizer, Inc., 153 S.W.3d 758, 764 (Ky. 2004) (citations omitted).  See Perez v. Wyeth Laboratories, Inc., 734 A.2d 1245, 1255 (N.J. 1999) (“courts do not wish to intrude upon the doctor-patient relationship”; “warnings that contradict information supplied by the physician will undermine the patient’s trust in the physician’s judgment”) (citation and quotation marks omitted).

The same is true for the learned intermediary rule in cases where plaintiffs seek to impose general duties on pharmacists to provide information direction to patients.  Several state high courts have made this crystal clear, the first being Pennsylvania (in a decision Bexis briefed). 

While the patient is entitled to know, and a doctor has a duty to inform the patient, of any dangers or side effects associated with a drug recommended for treatment, we see no sound reason for imposing on pharmacists the duty to supply information about the risks of drugs that have already been prescribed.  On the contrary, such a rule would have the effect of undermining the physician-patient relationship by engendering fear, doubt, and second-guessing.

Coyle v. Richardson-Merrell, Inc., 584 A.2d 1383, 1386 (Pa. 1991).  Massachusetts was next in line:

Physicians, after considering the history and needs of their patients and the qualities of the drug, are required to inform their patients of those side effects they determine are necessary and relevant for patients to know in making an informed decision.  Requiring the manufacturer to provide warnings directly to the consumer would interfere with the doctor-patient relationship.

Cottam v. CVS Pharmacy, 764 N.E.2d 814, 820 (Mass. 2002) (citations omitted).  Other states have followed:

Because the decision to prescribe a specific drug involves an analysis of the patient’s unique condition and a balancing of the risks and benefits of a given drug, the cases extending the learned intermediary to pharmacists reason that imposing a duty to warn on the pharmacist would intrude on the doctor-patient relationship and would force the pharmacist to practice medicine without a license.

Walls v. Alpharma USPD, Inc., 887 So.2d 881, 885 (Ala. 2004) (citation omitted).  Accord Kowalski v. Rose Drugs of Dardanelle, Inc., 378 S.W.3d 109, 120 (Ark. 2011) (“the underlying rationale for applying the learned-intermediary doctrine is to protect the patient-physician relationship by preventing pharmacists from second-guessing the physicians or otherwise interfering with the patient-physician relationship”); Nail v. Publix Super Markets, Inc., 72 So.3d 608, 615 (Ala. 2011) (the learned intermediary rule “sought to prevent, asking the pharmacist to intrude himself or herself into the physician-patient relationship”).

In addition to these high court decisions, numerous other state and federal courts have also recognized preservation of the physician-patient relationship as a key policy behind the learned intermediary rule.  The reasoning in Swayze v. McNeil Laboratories, Inc., 807 F.2d 464 (5th Cir. 1987) (applying Mississippi law), rejecting abandonment of the learned intermediary rule because doctors, allegedly, were not doing their jobs, remains valid today and almost universally applicable:

[W]hat would patients do with this information [a direct to patient warning]?  It might prompt would-be patients to put pressure on their physicians to better supervise, but physicians hardly need more incentive to be non-negligent.  In all likelihood, such warnings would only lead to confusion, and perhaps undermine the physician-patient relationship.  When the physician-patient relationship does exist, as here, we hesitate to encourage, much less require, a drug manufacturer to intervene in it. . . .  A special relationship, between physician and patient . . . receives special protection in law, and, at the same time, creates a great responsibility for every physician. . . .  [I]t is the physicians who have undertaken the responsibility of [performing the relevant medical procedures], and that responsibility cannot be shunted onto, or shared with, drug manufacturers.

Id. at 471.  Many other courts, applying the laws of many other jurisdictions, have recognized the essential role that the learned intermediary rule is intended to play in preserving the physician-patient relationship.

Alabama:  Lansdell v. American Home Products Corp., 1999 WL 33548541, at *5 (N.D. Ala. Oct. 26, 1999), held:

Neither manufacturer nor pharmacist has the medical education or knowledge of the medical history of the patient which would justify a judicial imposition of a duty to intrude into the physician-patient relationship.  Moreover, to hold otherwise, would require the pharmacist to question the physician’s judgment regarding the appropriateness of each customer’s prescription.  Such a policy, inevitably, would wedge the pharmacist into the relationship between a physician and her or his patient and, thereby, interfere with ongoing treatment.

Id. at *5 (citations and quotation marks omitted).  Accord Orr v. Wyeth-Ayerst Laboratories Co., 1999 WL 33548162, at *2 (Ala. Cir. Aug. 2, 1999) (direct-to-patient warnings are an “interference in the patient-physician relationship [that] can only do more harm than good).

Arkansas:  Bell v. Pfizer, Inc., 716 F.3d 1087, 1097 (8th Cir. 2013) (“[A] duty to warn the user directly would interfere with the relationship between the doctor and the patient.  With respect to the relationship between the patient, the physician, and the pharmaceutical manufacturer, the patient must look to the physician, for it is only the physician who can relate the propensities of the drug to the physical idiosyncracies of the patient.”) (citations and quotation marks omitted) (applying Arkansas law); Kendrick v. Wright Medical Technology, Inc., 2021 WL 3516663, at *6 (E.D. Ark. Aug. 10, 2021) (“imposing such a duty to warn on [the manufacturer] would interfere with the relationship between [plaintiff’s surgeon] and [plaintiff]”) (citation and quotation marks omitted); Stube v. Pfizer Inc., 446 F. Supp. 3d 424, 440 (W.D. Ark. 2020) (following West; “requiring drug manufacturers to directly warn the ultimate drug consumer would interfere with the doctor/patient relationship because patients rely on their doctors’ expertise in selecting and using drugs”); Kohl v. American Home Products Corp., 78 F. Supp. 2d 885, 890 (W.D. Ark. 1999) (courts “have been cautious in placing a duty to warn of potential side effects on a pharmacy because such a duty would interfere with the physician-patient relationship”).

California:  Plenger v. Alza Corp., 13 Cal. Rptr. 2d 811, 819 n.6 (Cal. App. 1992):

The rationale of the [learned intermediary] rule is:  [that] The doctor is intended to be an intervening party in the full sense of the word.  Medical ethics as well as medical practice dictate independent judgment, unaffected by the manufacturer’s control, on the part of the doctor. . . .  Were the patient to be given the complete and highly technical information on the adverse possibility associated with the use of the drug, he would have no way to evaluate it, and in his limited understanding he might actually object to the use of the drug, thereby jeopardizing his life.

(quoting Carmichael v. Reitz, 95 Cal. Rptr. 381, 400 (Cal. App. 1971)); Saavedra v. Eli Lilly & Co., 2013 WL 3148923, at *2 (C.D. Cal. June 13, 2013) (“Requiring the manufacturer to provide warnings directly to the consumer would interfere with the doctor-patient relationship.”) (applying California, Massachusetts and Missouri law); Regwan v. Reagan, 2020 WL 7862818, at *4 (Cal. Super. Nov. 5, 2020) (“Plaintiff’s allegations that a device manufacturer owes a duty to supervise and intervene in the provision of medical care cannot be reconciled with California’s licensing scheme or the learned intermediary doctrine and its underlying rationale.  To impose such a duty would place a medical device manufacturer in the middle of the doctor-patient relationship.”); Russo v. Safeway, Inc., 2006 WL 4114000 (Cal. Super. Jan. 3, 2006) (“[T]he rationale for applying the doctrine to pharmacists is compelling.  To impose such a duty to warn on the pharmacist would be to place the pharmacist in the middle of the doctor-patient relationship”.) (footnote omitted).

Connecticut:  DiVincenzo v. Molinaro, 2022 WL 17102332, at *2 (Conn. Super. Nov. 2, 2022) (“To impose a duty to warn on the pharmacist, however, would be to place the pharmacist between the physician who, having prescribed the drug presumably knows the patient’s present condition as well as his or her complete medical history, and the patient. Such interference in the patient-physician relationship can only do more harm than good.”) (citation and quotation marks omitted); Lynch v. State, 2018 WL 2207813, at *4 (Conn. Super. April 17, 2018). (analogizing sperm bank to pharmacist; “physicians have more knowledge than pharmacists about their patients’ needs and proclivities.  Holding otherwise would put pharmacists between the physician-patient relationship.”); Deed v. Walgreen Co., 927 A.2d 1001. 1004 (Conn. Super. 2007) (“To impose a duty to warn on the pharmacist would be to place the pharmacist in the middle of the doctor-patient relationship, without the physician’s knowledge of the patient.”) (citation and quotation marks omitted) (emphasis original).

Florida:  Buckner v. Allergan Pharmaceuticals, Inc., 400 So.2d 820, 823-24 (Fla. App. 1981) (purported physician failures to warn do not affect the learned intermediary rule; “extent of disclosure is a matter of medical judgment”; “failure to do so in a particular instance should not give rise to a duty in the manufacturer”); Pringle v. Johnson & Johnson, 2020 WL 4501834, at *3 n.5 (S.D. Fla. Jan. 30, 2020) (“[w]arnings directly from the manufacturer to the patient can undermine the doctor-patient relationship”); Labzda v. Purdue Pharma, L.P., 292 F. Supp.2d 1346, 1355 (S.D. Fla. 2003) (“Florida law does not impose a duty on the defendants to interfere with the physician-patient relationship”).

Georgia: Williams v. American Medical Systems, 548 S.E.2d 371, 375 (Ga. App. 2001) (“any information supplied to [plaintiff] by [the manufacturer] regarding dangers associated with the device could compromise the doctor-patient relationship”); Lance v. American Edwards Laboratories, 452 S.E.2d 185, 187 (Ga. App. 1994) (“any information supplied to [plaintiff] by defendants regarding dangers associated with use of the [device] may have compromised the doctor-patient relationship and thus impaired the best medical treatment for [plaintiff’s] chronic condition”); Walker v. Jack Eckerd Corp., 434 S.E.2d 63, 67 (Ga. App. 1993) (learned intermediary rule reflects “the need for preserving, without interference of third parties, a trusted physician-patient relationship”); Ellis v. C.R. Bard, Inc., 311 F.3d 1272, 1280 (11th Cir. 2002) (quoting and following Lance); Ingram v. Mylan Pharmaceuticals, Inc., 2009 WL 10665022, at *4 (N.D. Ga. Nov. 12, 2009). (“Considering the reality and practicality of the doctor and patient relationship, the learned intermediary doctrine developed in Georgia.”).

Illinois:  Urbaniak v. American Drug Stores, LLC, 126 N.E.3d 561, 567 (Ill. App. 2019) (“the learned intermediary doctrine dictates that pharmacists stay out of the physician-patient relationship”).  The same court made same point at greater length in Kennedy v. Medtronic, Inc., 851 N.E.2d 778 (Ill. App. 2006):

[A] central aspect of the learned intermediary doctrine . . . is that a licensed physician . . . has the knowledge of his patient’s medical history and background, and, therefore, he is in a better position, utilizing his medical judgment, to determine a patient’s needs and what medical care should be provided.  It would be unreasonable, and potentially harmful, to require a [device manufacturer’s representative] to delay or prevent a medical procedure simply because she believes the setting is not appropriate or the doctor is unqualified.  To hold otherwise would place a medical device manufacturer . . . in the middle of the doctor-patient relationship.

Id. at 786.  The same reasoning was also made, as to a pharmacist, in Fakhouri v. Taylor, 618 N.E.2d 518 (Ill. App. 1993):

Determining which medication is to be utilized in any given case requires an individualized medical judgment, which, in our opinion, only the patient’s physician can provide.  That physician, having prescribed the drug, presumably knows the patient’s current condition, as well as the patient’s complete medical history.  To impose a duty to warn on the pharmacist would be to place the pharmacist in the middle of the doctor-patient relationship, without the physician’s knowledge of the patient.

Id. at 521; accord Eldridge v. Eli Lilly & Co., 485 N.E.2d 551, 553 (Ill. App. 1985) (without the learned intermediary rule, “the pharmacist would have to interject himself into the doctor-patient relationship and practice medicine without a license”); In re Depo-Provera (Depot Medroxyprogesterone Acetate) Products Liability Litigation, 2025 WL 2711592, at *3 (N.D. Fla. Sept. 22, 2025) (following Fakhouri) (applying Illinois law); In re Yasmin & Yaz (Drospirenone) Marketing, Sales Practices & Products Liability Litigation, 692 F. Supp.2d 1025, 1033 (S.D. Ill. 2010) (policy “reasons” supporting learned intermediary rule include avoiding “[i]nterference with the doctor-patient relationship”); In re Yasmin & Yaz (Drospirenone) Marketing, Sales Practices & Products Liability Litigation, 692 F. Supp.2d 1012, 1018 (S.D. Ill. 2010) (same); Stephens v. CVS Pharmacy, 2009 WL 1916402, at *3 (N.D. Ill. June 11, 2009) (following Fakhouri); Jones v. Irvin, 602 F. Supp. 399, 402 (S.D. Ill. 1985) (“Placing these duties to warn on the pharmacist would only serve to compel the pharmacist to second guess every prescription a doctor orders in an attempt to escape liability.”).

Indiana:  Ingram v. Hook’s Drugs, Inc., 476 N.E.2d 881 (Ind. App. 1985):

The decision of weighing the benefits of a medication against potential dangers that are associated with it requires an individualized medical judgment.  This individualized treatment is available in the context of a physician-patient relationship which has the benefits of medical history and extensive medical examinations. . . .  The injection of a third-party in the form of a pharmacist into the physician-patient relationship could undercut the effectiveness of the ongoing medical treatment.

Id. at 886-87 (footnote omitted).

Iowa: Madsen v. American Home Products Corp., 477 F. Supp.2d 1025, 1033 (E.D. Mo. 2007) (“states want to preserve the doctor-patient relationship which could be undermined if patients received warnings from drug manufacturers that differed from doctor’s warnings”) (applying Iowa law).

Kansas:  Nichols v. Central Merchandise, Inc., 817 P.2d 1131, 1133 (Kan. App. 1991) (imposing a duty to warn on the pharmacist would intrude on the doctor-patient relationship and would force the pharmacist to practice medicine without a license”).

Maine:  Tardy v. Eli Lilly & Co., 2004 WL 1925536, at *2 (Me. Super. Aug. 3, 2004) (“The injection of a third party . . . into the physician/patient relationship could undercut the effectiveness of the ongoing medical treatment by the physician. . . .  [T]he patient would be receiving information about the risks of medication . . . from someone unfamiliar with the patient’s medical condition, after those risks had already been weighed by a physician having specific knowledge of the patient’s medical needs.”).

Massachusetts:  Yout v. Biogen Idec, Inc., 2015 WL 13697200, at *4 (Mass. Super. Oct. 15, 2015) (“Requiring the manufacturer to provide warnings directly to the consumer would interfere with the doctor-patient relationship.”).

Michigan:  Mowery v. Crittenton Hospital, 400 N.W.2d 633, 637 (Mich. App. 1986) (“a duty to warn the user directly would cause undue interference with the relationship between doctor and patient”) (citations omitted); Nichols v. McNeilab, Inc., 850 F. Supp. 562, 570 (E.D. Mich. 1993) (“The rationale behind the [learned intermediary rule] has been stated as avoiding the confusion and unnecessary interference with the doctor/patient relationship which might result if drug manufacturers were required to warn patients directly of their drug’s possible side effects.”); Hickey v. Tennessee Gas Pipeline Co., 1992 WL 236915, at *3 (E.D. Pa. Sept. 9, 1992) (“A warning to the patient under these circumstances could potentially cause undue interference with the doctor-patient relationship, cause patient confusion, and result in a hampering of the healing process.”) (quoting and following Mowery) (applying Michigan law); Reaves v. Ortho Pharmaceutical Corp., 765 F. Supp. 1287, 1289 (E.D. Mich. 1991) (“[t]he learned intermediary doctrine protects the physician-patient relationship”).

Missouri:  Puricelli v. Genetech, Inc., 2011 WL 1576779, at *2 n.2 (E.D. Mo. April 26, 2011) (“states want to preserve the doctor-patient relationship which could be undermined if patients received warnings from drug manufacturers that differed from doctor’s warnings”) (citation and quotation marks omitted).

New Jersey:  Jones v. Synthes USA Sales, LLC, 2010 WL 3311840, at *4 n.3 (D.N.J. Aug. 19, 2010) (the learned intermediary rule has a rationale which . . . prevents the court from intruding upon the doctor-patient relationship”); Simineri v. LifeCell Corp., 2015 WL 3384588, at *8 (N.J. Super. Law Div. May 8, 2015) (“To require [defendant] to communicate directly with the patient would unnecessarily intrude on the doctor-patient relationship.).

New York:  In a pharmacy case, Abrams v. Bute, 27 N.Y.S.3d 58 (N.Y. App. Div. 2016), held:

Courts have been reluctant to impose a standard of care on pharmacists that would . . . require pharmacists to exercise their own professional judgment in a manner that could conflict with that of the prescribing physician. . . .  Accordingly, a rule that requires a pharmacist to independently evaluate the propriety of a physician’s prescription would not only place an undue burden on pharmacists, but would likely create antagonistic relations between pharmacists and physicians and interfere with the patient-physician relationship.

Id. at 66-67 (citations and quotation marks omitted).  Accord Brumaghim v. Eckel, 944 N.Y.S.2d 329, 332 (N.Y. App. Div. 2012)  (imposing “duties to warn on the pharmacist would only serve to compel the pharmacist to second guess every prescription a doctor orders in an attempt to escape liability”); Bee v. Novartis Pharmaceuticals Corp., 18 F. Supp.3d 268, 295 (E.D.N.Y. 2014) (quoting and following Hogan); Hogan v. Novartis Pharmaceuticals Corp., 2011 WL 1533467, at *10 (E.D.N.Y. April 24, 2011) (“Broadly speaking, the learned intermediary rule seeks to preserve the doctor-patient relationship and allows the doctor to interpret the dangers involved in taking a drug; a warning to the patient, the rationale suggests, even if practical, could be detrimental as the patient may not properly weigh the drug’s risks against its benefits.”).

Oklahoma:  Carista v. Valuck, 394 P.3d 253, 259 (Okla. App. 2016) (rejecting “broad expansion to the duties of pharmacists” that would “second-guess the judgment of a doctor in prescribing drugs” and “would result in pharmacists acting as a form of second medical opinion, and certainly could enmesh pharmacists in the practice of medicine.”  Rather pharmacist duties are “limited by the principles and exceptions to the learned intermediary doctrine”) (footnote omitted).

Pennsylvania:  Makripodis v. Merrell-Dow Pharmaceuticals, Inc., 523 A.2d 374, 378 (Pa. Super. 1987) (“A requirement that a pharmacist provide to the consumer the same warnings that the drug manufacturer must supply to prescribing physicians would intrude upon the doctor-patient relationship.”); Polt v. Sandoz, Inc., 462 F. Supp.3d 557, 570 & n.18 (E.D. Pa. 2020) (quoting and following Coyle; “the physician, not the manufacturer, decides whether to prescribe drugs and explains the risks to consumers”); Ramirez v. Richardson-Merrell, Inc., 628 F. Supp. 85, 88 (E.D. Pa. 1986) (“To impose a duty to warn on the pharmacist, however, would be to place the pharmacist between the physician who, having prescribed the drug presumably knows the patient’s present condition as well as his or her complete medical history, and the patient. Such interference in the patient-physician relationship can only do more harm than good.”).

Rhode Island: Franks v. Coopersurgical, Inc., 722 F. Supp.3d 63, 91 (D.R.I. 2024) (the learned intermediary rule “puts the onus on a patient’s physician to inform her of the risks associated with a particular medical device because the physician is in a better position to understand the risks, and inform the patient of those risks by virtue of their doctor-patient relationship”) (citation omitted).

South Carolina:  Brooks v. Medtronic, Inc., 750 F.2d 1227, 1232 (4th Cir. 1984) (“[T]otal disclosure by a manufacturer would not be in the patient’s best interest.  One in a serious medical condition faces unwanted, unsettling and potentially harmful risks if advice, almost inevitably involved and longwinded, from non-physicians, contrary to what the doctor of his choice has decided should be done, must be supplied to him during the already stressful period shortly before his trip to the operating room.”).

Tennessee:  Laws v. Johnson, 799 S.W.2d 249, 254 (Tenn. App. 1990) (quoting and following Brooks, see South Carolina); Dunkin v. Syntex Laboratories, Inc., 443 F. Supp. 121, 123 (W.D. Tenn. 1977) (“attempts to give detailed warnings to patients could mislead patients and might also tend to interfere with the physician/patient relationship”).

Texas:  Butler v. Juno Therapeutics, Inc., 541 F. Supp. 3d 774, 786 (S.D. Tex. 2021) (“Texas law does not tie the [learned intermediary rule’s] protection to the specific conduct of the physician.  Rather, the protection is triggered by the presence of a physician-patient relationship.”); McKay v. Novartis Pharmaceuticals Corp., 934 F. Supp.2d 898, 910 n.90 (W.D. Tex. 2013) (multiple warnings from different sources would “undermine the prescribing doctor’s physician-patient relationship”), aff’d, 751 F.3d 694 (5th Cir. 2014); Solomon v. Bristol-Myers Squibb Co., 916 F. Supp.2d 556, 563 (D.N.J. 2013) (“Recognizing the doctor-patient relationship, the [learned intermediary rule] excuses a drug manufacturer from warning each patient who receives the product when the manufacturer properly warns the prescribing physician of the product’s dangers.”) (citation and quotation marks omitted) (applying Texas law).

Vermont:  Baker v. University of Vermont, 2005 WL 6280644 (Vt. Super. May 4, 2005) (“[R]equiring the pharmacist to warn of potential risks associated with a drug would interject the pharmacist into the physician-patient relationship and interfere with ongoing treatment.”) (citation and quotation marks omitted).

Washington:  Long v. Rite Aid Headquarters Corp., 2019 WL 1370442, at *3 (Wash. App. (March 25, 2019) (“The duty to warn about potential adverse side effects must be the sole obligation of the prescribing physician because the physician may often have valid reasons for deviating from the drug manufacturer’s recommendations based on a patient’s unique condition.”; “excessive warnings . . . could cause unfounded fear and mistrust of the physician’s judgment, jeopardizing the physician-patient relationship and hindering treatment”) (quotation marks and footnotes omitted).

West Virginia:  City of Charleston v. Joint Commission, 473 F. Supp. 3d 596, 625 (S.D.W. Va. 2020) (The learned intermediary rule reflects both “the concern that direct warnings to ultimate users would interfere with doctor/patient relationships” and “the corollary principle that physicians exercise their independent clinical judgment in patient evaluations”) (citation and quotation marks omitted); Ashworth v. Albers Medical, Inc., 395 F. Supp.2d 395, 407 (S.D.W. Va. 2005) (the learned intermediary rule “arises from the recognition of the primary role of the patient’s physician in diagnosing a particular condition or ailment and prescribing a course of treatment and from the reluctance to place a duty on a manufacturer that may interfere with the physician-patient relationship”).

Based on this survey of the law – including decisions from 29 different jurisdictions as recent as 2024 and 2025, and as old as 1971 −  it should be unmistakable that protection of the physician-patient relationship and prevention of that relationship being undermined by overly expansive warning duties concerning prescription medical products are at the core of the universally applicable learned intermediary rule.  This is not some tangential concern that only governs a small subset of learned intermediary cases.  No.  Preservation of the physician-patient relationship from disruption by third-party sources of generalized medical information, be they manufacturers, distributors, pharmacists, or whatever, is at the heart of the learned intermediary rule.

That’s why we are disturbed by the rationale of the recent decision in Himes v. Somatics, LLC, 549 P.3d 916, 926 (Cal. 2024) (discussed here).  Himes embraced a pernicious theory of warning causation that is fundamentally at odds with the learned intermediary rule’s purpose of protecting the physician-patient relationship from damage by product liability litigation.  Here’s why – it allows plaintiffs to get to the jury with otherwise hopeless causation evidence by testifying that had they known of some prescription-product-related risk they would have rejected their physician’s recommendation to use the product in question.  Forget possibly divergent warnings from different sources, we can think of nothing more corrosive of the physician-patient relationship than the prospect of liability turning on patient non-compliance with medical recommendations.

Here’s how things happened in Himes:  The plaintiff had previously been “prescribed at least nine different antipsychotics and antidepressants to treat depression, but her condition worsened,” and plaintiff “was hospitalized several times for severe depression or suicidal ideation.”  549 P.3d at 921.  As a last resort, one of her physicians treated plaintiff with electroshock therapy using the defendant’s device.  Id.  She was warned that the therapy “could cause short-term memory loss” but claimed that the defendant’s label did not disclose “permanent” brain damage and amnesia.  Id.  The doctor testified, however, “that he still would have recommended [the therapy] even if he had been informed of th[ose] risk[s].”  Id. at 922.  That additional information would not have changed a prescribing physician’s treatment is a well-established basis for severing causation as a matter of law in learned intermediary rule cases.

But Himes held that the plaintiff could overcome this testimony with evidence that “the [prescribing] physician would have communicated the stronger warning” to the plaintiff, and further that “an objectively prudent person in the patient’s position would have thereafter declined the treatment notwithstanding the physician’s continued recommendation of the treatment.”  Id. at 921 (emphasis added).

That formulation is totally at odds with all the above precedent establishing that preservation of the relationship of trust represented by the physician-patient relationship is a fundamental goal of the learned intermediary rule.  Even though there was significant California law favoring the preservation of the physician-patient relationship (see above), neither that precedent (be it from California or elsewhere), nor the relationship itself was mentioned in Himes.  Instead, Himes analogized to informed consent. “Implicit in our informed consent rule is the recognition that patients will sometimes opt out of the medical treatments their physicians recommend, as is their right.”  Id. at 926.  We agree with that statement as far as it goes, but we don’t think that a patient’s right to act contrary to medical advice should be weaponized into a cause of action – nor has any other court that we know of.

Moreover, the sort of evidence that Himes envisioned is not just plaintiffs with dollar signs in their eyes testifying that “if they had only known,” of course they would never have consented to the recommended drug/device/vaccine.  We’re not talking about Taxotere plaintiffs claiming that they, personally, would have refused to take a life-saving cancer drug because of hair loss or crying excessive tears.  See Id. at 932 (“If a subjective test were used, a plaintiff could simply offer self-serving testimony asserting that he or she would have declined the recommended treatment after being informed of the risks.”).

No, this formulation is more pernicious than a plaintiff relying on his or her own obviously biased testimony.  Himes demands purportedly “objective” testimony.  This test potentially requires rummaging through the entire physician-patient relationship in search of a basis for a jury to conclude that the physician’s judgment was so incorrect that an “objectively prudent patient” would have refused to comply:

[R]elevant factors that should be considered in determining whether an objectively prudent patient would have declined physician-recommended treatment include, but are not necessarily limited to, whether the physician weighed and assessed the risks and benefits of the treatment and, after discussing those risks and benefits with the patient, continued to recommend the treatment; whether the treatment was novel or was instead an established method for addressing the patient’s condition; the availability and utility of alternative treatments and the degree to which they have previously been tried in an effort to address the patient’s condition; the severity of the patient’s condition; and the likelihood that the treatment would have resulted in more than marginal benefits to the patient.

Id. at 931.

Thus, Himes has authorized plaintiffs – and eventually, juries, to second-guess essentially every aspect of medical decision-making – the quality of the doctor’s initial prescription decision; the quality of the discussion between the physician and the patient; whether the treatment was “novel” or “established”; alternative treatments; and the extent of both the medical risks and benefits.  Basically, under the guise of proving “causation,” Himes has imported into California prescription product liability litigation the entire range of proof normally seen in informed consent cases where the plaintiff sues his or her physician.  The plaintiff is thus encouraged to attack every aspect of the physician’s decision-making.  Such proof is necessarily destructive of the physician-patient relationship and thus directly at odds with one of the nationally accepted pillars of the learned intermediary rule.

We also point out that Himes is fundamentally different from another adverse learned intermediary rule decision, Blackburn v. Shire U.S., Inc., 380 So.3d 354 (Ala. 2022) (discussed here), that we have often criticized in the same breath for dumbing down learned intermediary rule-related causation requirements.  While Blackburn did indeed liberalize warning causation in Alabama, it did so in a way that did not require attacks on the functioning of the physician-patient relationship.  Rather, Blackburn was dependent on the physician’s own conduct, as testified to by the physician.  Blackburn involved instructions on how to monitor the development of adverse reactions.  Thus, a major question was “whether information provided by a prescription-drug manufacturer would have changed how a physician chose to monitor a patient.”  Id. at 368 (emphasis original).  The physician testified in Blackburn that if the defendant’s warnings had included more monitoring, “he ‘would have followed those protocols.’”  Id. at  357.  The plaintiff in Blackburn was not incentivized to attack the prescriber’s actions.  We think Blackburn was wrongheaded for other reasons – such as ignoring the undisputed fact that the plaintiff failed to keep subsequent medical appointments, and thus could not have been monitored at all.  Blackburn v. Shire U.S., Inc., 2020 WL 2840089, at *5 (N.D. Al. June 1, 2020).  But as bad as Blackburn was, at least it did not encourage patients to claim they would have disregarded medical device as a means to prove causation.

Photo of Steven Boranian

Can states compel prescription drug manufacturers to deliver steeply discounted medicines to unlimited numbers of pharmacies?  That is the issue that has been kicking around federal courts for a few years now, and the Fourth Circuit has now weighed in by answering “no.” 

We are talking here about drugs purchased under the federal government’s 340B Drug Pricing Program, and the Fourth Circuit’s recent opinion is significant for a couple of reasons.  First, the stakes are extremely high: Eligible purchasers bought $81.4 billion in covered outpatient drugs under the 340B program in 2024, and that figure is rising rapidly year by year.  Second, the Fourth Circuit’s opinion creates a circuit split on an issue that has generated intense interest from healthcare providers and pharmaceutical manufacturers alike.  In other words, the odds of the Supreme Court stepping in have increased.

The case is Pharmaceutical Research and Manufacturers of America v. McCuskey, No. 25-1054, 2026 WL 898259 (4th Cir. Mar. 31, 2026) (to be published in F.4th), and the opinion addresses whether states can prevent pharmaceutical manufacturers from limiting delivery of 340B-discounted drugs to so-called “contract pharmacies.” 

In a nutshell, if drug manufacturers want to participate in Medicare and Medicaid, they have to offer their drugs at a discount to certain healthcare providers—called “covered entities”—which typically care for low-income and rural populations.  When Congress first created the 340B program in 1992, few covered entities had in-house pharmacies, so HHS issued a guidance allowing each covered entity to use one external contract pharmacy where the covered entity’s patients could fill their prescriptions.

But then, in a guidance issued concurrently with the passage of the Affordable Care Act in 2010, HHS said that covered entities could use an unlimited number of contract pharmacies, which caused the use of contract pharmacies to increase twentyfold.  This is a problem and increases the risk of abuse through duplicate discounts and drug diversion.  Several drug manufacturers therefore imposed their own pharmacy limits, and the Third Circuit and the DC Circuit have ruled that federal law allows manufacturers to implement those limits.  We reported on those cases here and here. 

States therefore have stepped in and passed their own laws prohibiting manufacturers from limiting the numbers of contract pharmacies, or even requesting that covered entities and pharmacies submit claims data supporting eligibility for the discounted prices.  Drug manufacturers predictably have challenged these laws in multiple courts (see our latest report on those cases here).  Most courts have ruled against the manufacturers, including the Fifth and Eighth Circuits.  See AbbVie, Inc. v. Murrill, 166 F.4th 528 (5th Cir. 2026); Pharmaceutical Research Mfrs. v. McClain, 95 F.4th 1136 (8th Cir. 2024). 

The Fourth Circuit has gone the other way.  Like other states, West Virginia passed a contract pharmacy law that prohibits manufacturers from limiting the acquisition of 340B-discounted drugs or requesting claims data to verify eligibility.  The Fourth Circuit held that federal law preempted West Virginia’s law because the state law specially targeted participants in the federal 340B program (drug manufacturers) and altered the bargain that Congress struck with manufacturers when it created the program. 

The key issue is this: While other courts have characterized state contract pharmacy laws as merely regulating delivery of medicines, which is purportedly outside the scope of the federal 340B statute, the Fourth Circuit recognized that West Virginia was really regulating price, which is what the federal 340B statute is all about.  As the Fourth Circuit put it,

West Virginia . . . imposes conditions on drug manufacturers by virtue of their participation in the 340B program.  The result is that more drugs will be sold at bargain prices.

McCuskey, 2026 WL 2026 WL 898259, at *1 (emphasis added).  Price versus delivery.  That’s probably the most simplistic way to see this controversy.  States can regulate health and safety, and they ostensibly can regulate how covered entities have discounted drugs delivered to pharmacies, and by extension to their patients. 

But does that really make a difference?  The Fourth Circuit didn’t think so.  Government programs grounded in Congress’s spending power operate based on consent.  For the 340B program, that is true in a constitutional sense and also literally—the manufacturers participate in the 340B program under a Pharmaceutical Pricing Agreement executed with HHS.  Contract pharmacy laws like West Virginia’s interfere with that agreement. 

For one thing, the state law facially targets a federal domain and “springs obligations on manufacturers specifically by virtue of their participation in a federal program.”  Id. at *7.  Moreover, in conceiving the 340B program, Congress leveraged its spending power to forge a bargain, offering federal funds (Medicare and Medicaid spending) in exchange for compliance with certain conditions (offering drugs to covered entities at steep discounts).  West Virginia’s law seeks to “add conditions, uninvited,” thus disrupting the bargain.  Id. at *7-*8. 

The result is federal preemption.  The law directly purported to change the terms of the federal 340B program, where Congress “did not merely set a floor to which States may add additional obligations.”  Id. at *9.  Approving West Virginia’s law would upset the bargain that Congress struck, and it would also allow states to enact additional and unique obligations whenever they are unhappy with federal requirements. 

There were more specific problems with the statute.  The law intruded on HHS’s enforcement authority in numerous ways.  And it likewise interfered with audits by prohibiting manufacturers from requesting claims data.  Thus, “[t]he substantive provisions interfere at a high level with Congress’s exercise of its spending power and at an operational level with HHS’s enforcement authority and specific enforcement activities.  No possible construction of [West Virginia’s law] avoids this result.  As such, it is likely preempted.”  Id. at *12.

The PhRMA v.McCuskey opinion does not stand alone.  As we were writing this blogpost, the Fourth Circuit filed an unpublished opinion holding that Maryland’s contract pharmacy law was also likely preempted.  See Abbvie, Inc. v. Brown, No. 24-1939, 2026 U.S. App. LEXIS 10581 (4th Cir. Apr. 14, 2026).  One judge dissented from both opinions, with a long dissent in McCuskey that echoed points made in the Fifth Circuit and Eighth Circuit opinions—mainly that the West Virginia statute “regulates delivery.”  We understand that point, but as noted above, the difference between regulating “delivery” and regulating “price” is difficult to articulate.  The pharmacies are purchasing the medicines no matter what.  The only question is what price they will pay for them.  Surely a cert petition is coming, if not filed already. 

Drug & Device Resources

Bexis’ Book

FDA-regulated products now account for an estimated one-fifth of overall economic activity in the U.S. They have also been the focus of a litigation explosion. This timely guide covers all aspects of litigation involving drugs, medical devices, vaccines and other FDA-regulated prescription products.
Learn more about this book